org.openimaj.image.model.patch

Class PatchClassificationModel<Q,T extends Image<Q,T>>

java.lang.Object

org.openimaj.image.model.patch.PatchClassificationModel<Q,T>

Type Parameters:

Q - Type of pixel

T - Type of Image

All Implemented Interfaces:

Serializable, Cloneable, ImageClassificationModel<T>, EstimatableModel<T,FImage>, Model<T,FImage>

Direct Known Subclasses:

MBFPatchClassificationModel

public abstract class PatchClassificationModel<Q,T extends Image<Q,T>>
extends Object
implements ImageClassificationModel<T>

An ImageClassificationModel based on the idea of determining the probability of a class of a pixel given the local patch of pixels surrounding the pixel in question. A sliding window of a given size is moved across the image (with overlap), and the contents of the window are analysed to determine the probability belonging to the pixel at the centre of the window.

Author:

Jonathon Hare (jsh2@ecs.soton.ac.uk)

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected int	patchHeight
protected int	patchWidth

Constructor Summary

Constructors

Constructor and Description
PatchClassificationModel(int patchWidth, int patchHeight) Construct with the given dimensions for the sampling patch.

Method Summary

Methods

Modifier and Type	Method and Description
FImage	classifyImage(T im) Classify the given image and return the corresponding probability map
abstract float	classifyPatch(T patch) Classify a patch, returning the probability of the central pixel belonging to the class.
abstract PatchClassificationModel<Q,T>	clone() Clone the model
boolean	estimate(List<? extends IndependentPair<T,FImage>> data) Estimates the model from the observations in the list of data.
protected abstract T[]	getArray(int length)
int	numItemsToEstimate()
FImage	predict(T data) Uses the model to predict dependent data from an independent value.

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface org.openimaj.image.model.ImageClassificationModel

learnModel

Field Detail

patchHeight

protected int patchHeight

patchWidth

protected int patchWidth

Constructor Detail

PatchClassificationModel

public PatchClassificationModel(int patchWidth,
                        int patchHeight)

Construct with the given dimensions for the sampling patch.

Parameters:

patchWidth - the width of the sampling patch

patchHeight - the height of the sampling patch

Method Detail

classifyPatch

public abstract float classifyPatch(T patch)

Classify a patch, returning the probability of the central pixel belonging to the class.

Parameters:

patch - the patch.

Returns:

the probability of the central pixel belonging to the class.

classifyImage

public FImage classifyImage(T im)

Description copied from interface: ImageClassificationModel

Classify the given image and return the corresponding probability map

Specified by:

classifyImage in interface ImageClassificationModel<T extends Image<Q,T>>

Parameters:

im - the image to classify

Returns:

the probability map

clone

public abstract PatchClassificationModel<Q,T> clone()

Description copied from interface: EstimatableModel

Clone the model

Specified by:

clone in interface ImageClassificationModel<T extends Image<Q,T>>

Specified by:

clone in interface EstimatableModel<T extends Image<Q,T>,FImage>

Overrides:

clone in class Object

Returns:

a cloned copy

getArray

protected abstract T[] getArray(int length)

estimate

public boolean estimate(List<? extends IndependentPair<T,FImage>> data)

Description copied from interface: EstimatableModel

Estimates the model from the observations in the list of data. The data must contain at least EstimatableModel.numItemsToEstimate() pairs of dependent and independent data. It may contain more, in which case the estimate method may choose to make use of this data for validation, or obtaining a better model by a least squares method for example.

Specified by:

estimate in interface EstimatableModel<T extends Image<Q,T>,FImage>

Parameters:

data - Data with which to estimate the model

Returns:

true if a successful estimate was made, false otherwise

See Also:

EstimatableModel.numItemsToEstimate()

numItemsToEstimate

public int numItemsToEstimate()

Specified by:

numItemsToEstimate in interface EstimatableModel<T extends Image<Q,T>,FImage>

Returns:

The minimum number of observations required to estimate the model.

predict

public FImage predict(T data)

Description copied from interface: Model

Uses the model to predict dependent data from an independent value.

Specified by:

predict in interface Model<T extends Image<Q,T>,FImage>

Parameters:

data - the data (independent variable)

Returns:

Dependent variable(s) predicted from the independent ones.